Dockerfile

FROM ghcr.io/ggml-org/llama.cpp:full

WORKDIR /app

# Python仮想環境を作るための最低限のものを入れる
RUN apt-get update && apt-get install -y --no-install-recommends \
    python3-pip python3-venv \
    && rm -rf /var/lib/apt/lists/*

# venv を作成
RUN python3 -m venv /opt/venv
ENV PATH="/opt/venv/bin:$PATH"

# venv にだけ pip パッケージを入れる
RUN pip install --no-cache-dir -U huggingface_hub

# モデルとmmprojをダウンロード
RUN python3 -c 'from huggingface_hub import hf_hub_download; \
    repo="unsloth/Qwen3.6-27B-GGUF"; \
    hf_hub_download(repo_id=repo, filename="Qwen3.6-27B-IQ4_NL.gguf", local_dir="/app"); \
    hf_hub_download(repo_id=repo, filename="mmproj-BF16.gguf", local_dir="/app")'

CMD ["--server", \
     "-m", "/app/Qwen3.6-27B-IQ4_NL.gguf", \
     "--mmproj", "/app/mmproj-BF16.gguf", \
     "--host", "0.0.0.0", \
     "--port", "7860", \
     "-t", "2", \
     "--cache-type-k", "q8_0", \
     "--cache-type-v", "iq4_nl", \
     "-c", "64000", \
     "-n", "38912"]